Understanding Socioenvironmental Contributors to Racial and Ethnic Disparities in Disability Among Older Americans

Abstract

Our understanding of the mechanisms through which racial/ethnic disparities in disability in older adults develop and are maintained is limited. We examined the role of physical impairment, socioeconomic factors and health for racial/ethnic disparities in activities of daily living (ADL), and the modifying role of the indoor home environment. Data come from the National Health and Aging Trends Study (N = 5,640), and negative binomial regression models were specified separately for men and women. Blacks and Hispanics reported more ADL difficulty than Whites. Living in homes with clutter was associated with higher rates of ADL difficulty, but it was not related to racial/ethnic disparities. Racial/ethnic differences were explained by physical impairment for men, but not for women. Socioeconomic factors and health accounted for remaining disparities for Black, but not for Hispanic women. Attention to individual and environmental factors is necessary to fully understand and address race/ethnic disparities in disability in older Americans.

Keywords

disability disablement process home environment racial/ethnic disparities activities of daily living older age

At least 30% of Americans over the age of 65 have some degree of physical disability (Wan & Larsen, 2014), defined as a substantial limitation in life activities (Erickson, Lee, & von Schrader, 2010), which can result in dependency, institutionalization, and high rates of health-care utilization (Fried & Guralnik, 1997; Metzelthin et al., 2013). Researchers have also documented large racial/ethnic disparities in disability, with minorities experiencing higher rates of and more severe disability than Whites (Dunlop, Song, Manheim, Daviglus, & Chang, 2007; Mendes de Leon, Barnes, Bienias, Skarupski, & Evans, 2005; Schoeni, Martin, Andreski, & Freedman, 2005). Although the prevalence of disability has declined for many Americans over the past two decades (Seeman, Merkin, Crimmins, & Karlamangla, 2010), the non-Hispanic White/minority gap in disability has not narrowed (Schoeni et al., 2005). Despite these large racial/ethnic disparities in disability, our understanding of the mechanisms through which they develop and are maintained is limited.

Racial/ethnic disparities in disability are a major public health concern. Disparities in disability result in disproportionate economic and social strains (e.g., caregiving burden) on minority families and communities (Pinquart & Sörensen, 2005) and thus amplify existing financial and social inequities, which may contribute to continuing health disparities. Second, as the prevalence of chronic conditions and obesity increases, coupled with aging of the population, disability will become an increasingly consequential driver of racial/ethnic health disparities and racial/ethnic differences in mortality (Harper, 2014). In this article, we will first examine racial/ethnic differences in self-care disability in a nationally representative sample of Medicare eligible U.S. adults. Next, we will investigate factors that contribute to the racial/ethnic patterning in disability, including the home environment as a critical context in which adults interact every day.

Literature Review

Numerous studies have tried to identify the factors driving the race/ethnic differences in disability, focusing primarily on individual factors such as socioeconomic status (SES), health behaviors, and chronic conditions (Bowen, 2009; Kington & Smith, 1997; Latham, 2012, 2014; Thorpe et al., 2011). SES has been investigated as the primary causal mechanism through which racial/ethnic disparities in disability are generated (Haas & Rohlfsen, 2010; Kelley-Moore & Ferraro, 2004). Although SES is strongly associated with gradients in disability at all levels of income (Minkler, Fuller-Thomson, & Guralnik, 2006), and may account for differences in physical functioning (e.g., balance, ability to walk a certain distance), it does not fully account for racial/ethnic disparities in disability (Bowen, 2009; Fuller-Thomson, Nuru-Jeter, Minkler, & Guralnik, 2009).

Researchers have also examined the contribution of health behaviors and chronic conditions to racial/ethnic disparities in disability. Behavioral risk factors such as smoking, physical activity, and body mass index are established predictors of disability (Fried & Guralnik, 1997), but there is little evidence that they account for the racial/ethnic disparities in disability (Haas, Krueger, & Rohlfsen, 2012; Latham, 2014). Given the large disparities in disability that are not fully explained by individual socioeconomic factors, health status, or health behaviors, research that further probes the factors driving the racial/ethnic patterning in disability is warranted. Unmeasured factors at the socioenvironmental level that to date have not been thoroughly studied may contribute to race/ethnic disparities in disability. These unmeasured factors may also help explain the gender differences in associations between SES and physical health and race/ethnic disparities in disability that have been found previously (Mendes de Leon et al., 1997; Warner & Brown, 2011; Zsembik, Peek, & Peek, 2000).

Models of disablement, such as the Disablement Process (Verbrugge & Jette, 1994), the Nagi Framework (Nagi, 1965), and the International Classification of Disability, Functioning and Health (ICF; Freedman, 2009; World Health Organization, 2002), emphasize the complex interplay between environmental and individual factors in the pathway between pathology, physical impairment, functional limitations, and disability. These models define disability as a function of physical functioning (also referred to as physical impairment) and health conditions, as well as the physical and social environment, and factors intrinsic to the individual like behaviors and psychological factors (Freedman, 2009). Thus, research addressing the causes of racial/ethnic disparities in disability should consider the interplay of functional capacity, personal factors, and the environment.

The significance of the physical and social environment to the well-being and functioning of elderly adults has been well established in environmental gerontology (Wahl, Fänge, Oswald, Gitlin, & Iwarsson, 2009; Wahl, Schilling, Oswald, & Iwarsson, 2009) based on the environmental or competence press model, which links excessive environmental demands with impairment and disability (Wahl, Fänge, et al., 2009). The home is the most immediate physical environment in which basic self-care activities are carried out. Disablement models identify the environment, including the home environment, as an important context for intervention to reduce disability (Wahl, Fänge, et al., 2009). Research indicates, for example, that most falls occur inside the home (Nachreiner, Findorff, Wyman, & McCarthy, 2007), which supports the home as an important context in which the physical functioning of older adults should be addressed.

Theories of aging indicate that people spend more time in their homes as they grow older due to physical barriers, social ties to their homes, and the desire to age in place (Iwarsson et al., 2007). For the aging and disabled population, however, the home environment can also present a significant barrier to independence and safety (Gitlin, 2003; Wahl, Fänge, et al., 2009) as well as increased difficulty with physical functioning and self-care (Iwarsson, Isacsson, & Lanke, 1998; Gitlin, Mann, Machiko & Sue, 2001 ; Wahl, Schilling, et al., 2009). Barriers in the home may be particularly salient for individuals with physical or cognitive limitations, who are more vulnerable to their environments (Iwarsson et al., 1998; Wahl, Schilling, et al., 2009). For example, researchers found that individuals with more person–environment fit problems (generated by either physical limitations or increasing environmental challenges) were then more vulnerable to their environment (Iwarsson et al., 1998).

Although the home environment is not likely to cause disability, in accordance with models of disablement, a hazardous home environment could create challenges to meeting self-care and functioning needs for older individuals with physical impairments (Iwarsson et al., 2007). Disrepair inside the home including tripping hazards such as broken or uneven flooring and objects or clutter in walking paths may be particularly challenging for adults who have difficulty walking or performing other physical tasks (Iwarsson et al., 1998; Wahl, Fänge, et al., 2009). Disablement models such as the ICF distinguish between an individuals’ performance (actual functioning) in their current environment and their capacity, which is their highest level of functioning in any given moment (World Health Organization, 2002). This suggests that a person’s actual functioning is dependent on his or her current environment and that it is possible to improve a person’s capacity by making improvements to the home environment that facilitate functioning (e.g., installing grab bars in the bathtub). In this research, we will investigate whether physical functioning—measured by an objective performance-based measure—and the home environment interact to influence difficulty with self-care.

There is evidence that improving home environments is associated with higher functional ability outcomes (e.g., self-care; Szanton et al., 2015; Wahl, Fänge, et al., 2009). Differences in the home environment and associations between the home environment and functioning by gender and race/ethnicity have not been well studied, but researchers found that elderly minorities and women were more likely to encounter problems in their home environment (Gitlin, et al., 2001). Coupled with the documented lower levels of physical capacity and greater disability for non-White elderly adults (Fuller-Thomson et al., 2009; Haas et al., 2012; Latham, 2012; Thorpe Jr et al., 2008), the home environment could potentially be differentially associated with disability by race/ethnicity.

Few researchers have considered the complex interplay between multiple factors in the disablement process in the context of racial/ethnic disparities in disability, and no one has explored the influence of the home environment on racial/ethnic disparities in disability. Based on theory that posits that physical functioning/impairment interacts with the environment to influence functioning and disability (World Health Organization, 2002), as well as environmental press models and person–environment fit, we test interactions between physical impairment and the home environment on racial/ethnic differences in disability. For example, while clutter on the floor may not be associated with difficulty in self-care for an individual who is strong and able to walk well, clutter may make it challenging for someone with poor balance to get to the bathroom.

We also test the interaction between impairment and race/ethnicity on disability to determine whether the association between race/ethnicity and disability may depend on the degree of an individuals’ physical functioning. Although race/ethnicity is often conceptualized as a demographic factor, we apply a social perspective and consider the influence of race/ethnicity on social and environmental exposures across the life course that likely influence racial and ethnic differences in disability (Williams, Priest, & Anderson, 2016). Theoretically driven research with attention to the both individual and environmental factors is needed to fully understand the sources of racial/ethnic disparities in physical disability among older adults in the United States.

In this research, we strive to increase the understanding of the factors driving racial/ethnic disparities in disability by integrating individual measures of functioning, as defined by physical impairment, and disability as defined by self-care difficulty with measures of the home environment. The extent to which home environments vary by race/ethnicity (Gitlin, et al., 2001; Tabbarah, Mihelic, & Crimmins, 2001) may account for the unexplained differences in disability prevalence by race/ethnicity. For example, researchers found that 20% of older Blacks compared to 14% of older Whites reported discrepancies between their current home environment and their needs (Tabbarah et al., 2001). Moreover, we use a measure of disability that more closely maps to the activities that take place within the home environment and is reflective of interactions between the levels of functioning and the social and physical environment. At a fundamental level, basic self-care is an essential requirement for well-being and survival. Yet, the structure of the home environment may lead to differences in the levels of independence in daily activities. A recent intervention aimed at reducing home environmental barriers for older, disabled adults succeeded at reducing difficulty in activities of daily livings (ADLs) in almost 80% of the enrolled adults (Szanton et al., 2015). The brief and inexpensive intervention included simple home improvements to achieve functional goals. Results from this project suggest that even relatively basic, inexpensive repairs in the home environment can result in substantial improvements in functioning among older adults.

The objectives of this research are to (1) examine the racial/ethnic patterning of disability in a nationally representative sample of older Americans and (2) explore the association of physical, socioeconomic, and home environmental factors with disability, including the extent to which they explain any racial/ethnic patterning of disability. We examine these models separately for men and women based on the evidence that suggests that the risk factors and consequences of disability vary by gender (Penning & Strain, 1994). By using a standardized performance-based battery of physical capacity (reversed coded to represent physical impairment), we are able to better isolate their actual capacity (vs. performance) and test the interaction between capacity and their current home environment. Given the documented racial/ethnic disparities in physical impairment and the relationship between physical impairment and disability, we expect that physical impairment will largely account for any disparity in disability, although these relationships may vary by gender. We also hypothesize that socioeconomic factors and the indoor home environment, including interactions between physical impairment and the home environment, will explain some of the remaining racial/ethnic differences in disability for men and women.

Research Design

Data and Measures

Data come from the National Health and Aging Trends Study (Johns Hopkins University, NHATS Public Use Data, 2012), a nationally representative longitudinal study of the health and well-being of 8,245 Medicare recipients aged 65 and older. The study procedure was approved by the Johns Hopkins Bloomberg School of Public Health’s institutional review board. Black and the oldest adults were oversampled. Details of the sampling strategy and design are found elsewhere (Montaquila, Freedman, & Kasper, 2012). NHATS was designed to assess disability within a broad framework that includes the family, home and neighborhood environment, and society more generally. In-person interviews were used to collect data on disability, physical capacity, and sociodemographics. Information related to the home environment was collected via interviewer observations. The first round of NHATS occurred in 2011, with annual follow-ups. For this study, we used data from the second round of Johns Hopkins University, NHATS (2012), which included interviewer observations of conditions inside and outside participants’ homes. The response rate for Round 2, conditional on Round 1, was 85.6% for participants who completed a sample person interview. In our analysis, we focus on the 5,640 community-living participants, excluding 473 in nursing homes or assisted living.

Disability was measured by self-reported difficulty with ADL. Our ADL measure was adapted from Freedman et al. (2011) and includes six activities: eating, dressing, toileting, bathing/washing, getting around inside, and transferring. A dichotomous indicator of disability was created based on a previous framework that categorizes individuals into five disability groups including fully able, successful accommodation, reductions in activities, difficulty, and needs assistance. “Fully able” and “successful accommodation” were collapsed to represent independence in daily activities with or without using devices, while “reduced activities,” “difficulty,” and “assistance” were collapsed to represent at least some degree of difficulty or inability to complete daily activities without assistance (scored as 0/1 on a binary indicator for each of the ADL activities, respectively). The 6 ADL items were then summed to create a final ADL score for each individual ranging from 0 to 6, where higher values denote more disability.

Physical impairment was objectively assessed using the short physical performance battery (SPPB; Vasunilashorn et al., 2009), a reliable measure of lower extremity functioning that is predictive of mobility, disability, and mortality. The SPPB assesses gait speed over a 3-m walk, standing balance, and time to rise rapidly from a chair. NHATS used an established scoring approach, which is described in more detail elsewhere (Guralnik et al., 1994). The physical capacity score ranges from 0 to 12, with higher scores reflecting higher functioning. In a validation study of the SPPB (Vasunilashorn et al., 2009), a score of 10 was found to increase the risk of developing mobility difficulty 4-fold, and a score of 7 or less was found to increase the risk of incident mobility difficulty by 32 times that of an individual with the highest SPPB score. For our purposes, we created a physical impairment measure by reverse coding the SPPB score where higher values correspond to more impairment. We use the suggested clinically relevant levels of physical capacity (reverse coded) for the purposes of graphical presentation in our results. Approximately 13% of the analytic sample contained missing data on SPPB, so we used multiple imputations (MIs) to impute SPPB scores.

Educational attainment was assessed in Round 1 by nine categories ranging from “no schooling completed” to “master’s, professional, or doctoral,” and the categories were collapsed into less than high school, high school graduate/equivalent, and beyond high school. Marital status was collapsed into married/living with partner (reference category), separated/divorced, widowed, and never married. Total family income from all income and assets was assessed at baseline. Participants who were not able to give a dollar amount for total income were prompted using five bracketed ranges of total income from which to choose (13% of the sample). NHATS conducted an imputation using cylindrical hot deck imputation methods to create five imputed values for total income (31% of the sample; Montaquila et al., 2012). We computed a single income value for participants with missing income information by taking the median of the five imputed income values. Due to a highly skewed distribution, a natural log transformation of income was used in analyses.

A measure of disrepair inside the home was created by taking a sum of several interviewer observations including the presence of flooring disrepair, broken furniture, and trip hazards. The indoor disrepair measure ranges from 0 to 3 with higher values corresponding to more disrepair. Clutter was also assessed by interviewer observation of the room in which the interview was conducted and in other rooms in the home. An average of the two items was taken, and responses were collapsed into no clutter (no on both items) or clutter.

Due to the sensitive nature of birthdate information in NHATS, age at the time of the interview was only available in six 5-year age categories, which were collapsed into 65–74, 75–84, and 85+. Race/ethnicity was self-reported at baseline on primary race and ethnicity and collapsed into White, not Hispanic; Black, not Hispanic; Hispanic; and Other, not Hispanic. The race category “Other” includes individuals who identified as American Indian, Asian, Native Hawaiian, Pacific Islander, or another race that was not categorized. Chronic disease was included as a control for multiple comorbid conditions. Participants were asked to report whether a doctor diagnosed them with any of nine chronic conditions in the past year: heart disease, high blood pressure, arthritis, osteoporosis, diabetes, lung disease, stroke, dementia or Alzheimer’s disease, and cancer. A sum of chronic diseases was created, ranging from 0 to 8.

Statistical Analysis

Due to well-documented gender differences in disability (Murtagh & Hubert, 2004), we conducted gender-stratified analyses of racial/ethnic disparities in ADL. ADL difficulty ranged from 0 to 6. The count data were overdispersed (mean = 0.97, variance = 2.58), and the overdispersion parameter (α = 0.71) was significantly greater than zero (χ² for test = 9,611.00, p < .01), so we used negative binomial regression models to examine racial/ethnic disparities in the number of ADL difficulties. All coefficients and 95% confidence intervals (CIs) were exponentiated and reported as rate ratios (RRs). First, we determined whether race/ethnicity was associated with ADL difficulty separately for men and women (Model 1). Next, we added physical impairment to determine the contribution of impairment to any racial/ethnic disparities present in ADL difficulty (Model 2). Demographic, socioeconomic, and health factors were then included (Model 3), and the interactions between physical impairment and the indoor home environment were added (Model 4). Finally, we added an interaction between physical impairment and race/ethnicity (Model 5) to determine whether the effects of physical impairment on ADL difficulty varied by race/ethnicity due to other unmeasured factors varying by race.

All models included participants with nonmissing data on ADL difficulty and all covariates, which is over 90% of the community-living participants resulting in 2,059 men and 2,835 women with complete data. Participants dropped from the analysis were more likely to be male and non-White. All analyses were adjusted for sampling weights (Montaquila, Freedman, Spillman, & Kasper, 2014) using survey estimation commands in Stata (StataCorp, 2015) to account for unequal probabilities of selection. Due to the substantial amount of missing data for the SPPB, we performed a MI. A total of 15 multivariate MIs were performed using chained equations (White, Royston, & Wood, 2011). Predictive mean matching was used to impute physical capacity (SPPB) to adequately account for the truncated distributions (White et al., 2011), and all analyses were run on the 15 imputed data sets using MI estimation commands in Stata to adjust standard errors.¹

Results

Study sample characteristics are presented in Table 1. On average, participants had a relatively low level of difficulty with ADL (mean = 0.77), and women reported more difficulty than men (p < .001, Table 1). Average physical impairment for the full sample fell below the recommended clinical level corresponding to an increased risk of mobility disability (mean = 3.75), with women experiencing more impairment than men (p < .001). Women had lower income, were less educated, less likely to be married, and had more chronic diseases than men. Approximately 11% of the sample had homes characterized as having indoor disrepair and 27% of the homes contained at least some clutter, with women living in homes with more disrepair than men. The association between race and ADL difficulty varied by gender, so results are reported separately for men and women (p value for interaction <.01).

Table 1.

Mean (Standard Deviation) or Percentage of Sample Characteristics for the Full Sample and for Men and Women Separately (N = 5,640^a).^b

Variable	Total	Men (n = 2,401)	Women (n = 3,239)	p Value^c
ADL difficulty (Range 0–6)	0.77 (0.02)	0.66 (0.03)	0.86 (0.03)	<.001
Male (%)	2,401 (43.77)
Race/ethnicity (n, %)
White	3,851 (80.54)	1,677 (80.99)	2,174 (80.19)	>.05
Black	1,247 (8.39)	479 (7.54)	768 (9.07)
Hispanic	337 (6.68)	149 (6.84)	188 (6.54)
Other^d	205 (4.39)	96 (4.62)	109 (4.20)
Age (%)
65–74	2,094 (50.70)	970 (53.67)	1,124 (48.34)	<.05
75–84	2,350 (36.59)	1,006 (36.14)	1,344 (36.95)
85+	1,196 (12.70)	425 (10.19)	771 (14.71)
Marital status (%)				<.05
Married	2,838 (57.72)	1,697 (74.79)	1,141 (44.14)
Divorced	683 (12.28)	256 (10.17)	427 (13.97)
Widowed	1,913 (26.61)	360 (11.64)	1,553 (38.52)
Never married	205 (3.38)	87 (3.40)	118 (3.37)
Physical impairment (0–12)	3.75 (0.07)	3.24 (0.09)	4.16 (0.09)	<.05
Income (US$0–5,000,000)^e	54,113.92 (3,210.53)	68,025.26 (6,147.43)	43,047.39 (3,074.87)	<.05
Education (%)
<HS	1,472 (21.10)	610 (20.91)	862 (21.24)	<.05
HS	2,865 (53.29)	1,076 (46.54)	1,789 (58.66)
≥College	1,257 (25.61)	696 (32.54)	561 (20.10)
Indoor disrepair present (%)	715 (11.39)	289 (10.82)	426 (11.82)	<.05
Clutter present (%)	1,594 (26.77)	667 (26.70)	927 (26.82)	>.05
Total number of chronic diseases (0–9)	2.20 (0.02)	1.95 (0.03)	2.40 (0.03)	<.05

Note. ADL = activities of daily living; HS = high school.

^aMaximum sample size; variables contain missing values. ^bMeans/proportions were generated using survey weights to be generalizable to the U.S. population, sample size reflect actual sample. ^c p Value for male/female difference in means or proportions. ^dOther race includes American Indian/Asian/Native Hawaiian/Pacific Islander. ^eContinuous income is highly skewed; natural log of income used in all models.

Men

First, we examined racial/ethnic differences in disability (Table 2, Model 1). Black and Hispanic men experienced higher rates of ADL difficulty than White men (RR Black men = 1.55, 95% CI [1.22, 2.39]; RR Hispanic men = 1.59, 95% CI [1.05, 1.47]). The disparities were fully explained by differences in physical impairment (Model 2); a one-unit increase in impairment was associated with a 28% higher risk of difficulty performing daily activities (95% CI [1.25, 1.31]). Higher income was protective against ADL difficulty in all men, and higher chronic disease burden was associated with more ADL difficulty (Model 3, Table 2). Model 4 includes an interaction term between the home environment and physical impairment to determine whether the indoor home environment moderated the association between physical impairment and disability. The interaction term between indoor disrepair and impairment was not significant, so we omitted it from the model. The presence of clutter inside the home moderated the association of physical impairment on ADL difficulty, but men who lived in a home with clutter had higher rates of ADL difficulty at all levels of impairment, which is also indicative of a significant main effect of clutter on ADL difficulty (RR = 1.68, 95% CI [1.21, 2.34] for zero impairment). As the level of physical impairment increased, the presence of clutter had an even larger effect on disability (Figure 1).

Table 2.

Rate Ratios for Racial and Ethnic Disparities in Difficulty With Activities of Daily Living as a Function of Individual Factors and the Home Environment, Men (n = 2,059): U.S. National Health and Aging Trends Study (2012).^a

Variable	Model 1: Racial/Ethnic Differences in Disability	Model 2: Physical Impairment (PI)	Model 3: Socioeconomic and Health Factors	Model 4: Indoor Home Environment	Model 5: Impairment and Race/Ethnicity Interactions
Race/ethnicity^b
Black	1.55 (1.22, 2.39)**	1.03 (0.83, 1.26)	0.93 (0.76, 1.13)	0.91 (0.75, 1.11)	0.98 (0.67, 1.44)
Hispanic	1.59 (1.05, 1.47)*	1.02 (0.72, 1.46)	0.78 (0.53, 1.16)	0.80 (0.56, 1.15)	0.56 (0.29, 1.09)
Other	0.93 (0.59, 1.43)	0.95 (0.58, 1.57)	0.74 (0.44, 1.22)	0.73 (0.44, 1.20)	0.98 (0.45, 2.13)
Age^c
75 to 84	1.51 (1.22, 1.87)**	1.00 (0.83, 1.20)	0.98 (0.81, 1.18)	0.99 (0.82, 1.19)	0.99 (0.82, 1.19)
85+	2.82 (2.20, 3.60)**	1.20 (0.97, 1.49)	1.29 (1.01, 1.64)*	1.30 (1.03, 1.66)*	1.30 (1.02, 1.65)*
Marital status^d
Divorced			0.81 (0.57, 1.17)	0.79 (0.56, 1.12)	0.80 (0.56, 1.13)
Widowed			0.75 (0.58, 0.97)*	0.74 (0.57, 0.96)*	0.72 (0.55, 0.94)*
Never married			0.81 (0.50, 1.31)	0.79 (0.48, 1.29)	0.79 (0.48, 1.28)
PI (SPPB)		1.28 (1.25, 1.31)**	1.24 (1.21, 1.27)**	1.26 (1.22, 1.30)**	1.24 (1.20, 1.27)**
Education^e
High school			0.83 (0.67, 1.02)	0.84 (0.68, 1.03)	0.82 (0.67, 1.01)
≥College			1.01 (0.79, 1.31)	1.04 (0.80, 1.34)	1.01 (0.78, 1.31)
Income (ln)			0.78 (0.70, 0.88)**	0.79 (0.71, 0.89)**	0.79 (0.71, 0.89)**
Chronic disease			1.28 (1.21, 1.35)**	1.28 (1.21, 1.36)**	1.27 (1.20, 1.35)**
Indoor disrepair present				0.93 (0.74, 1.15)	0.91 (0.72, 1.14)
Indoor clutter present				1.68 (1.21, 2.34)**	1.29 (1.06,1.57)*
PI × Indoor Clutter				0.95 (0.91, 0.99)*
PI × Race/Ethnicity
PI × Black					0.99 (0.94, 1.04)
PI × Hispanic					1.05 (0.97, 1.14)
PI × Other					0.94 (0.86, 1.03)

Note. SPPB = short physical performance battery; HS = high school.

^aRate ratios derived from exponentiating the coefficients from negative binomial regression models. ^bReference group = White. ^cReference group = Age 65–74. ^dReference group = married/living with partner. ^eReference group = less than high school education.

Significant at *p < 0.05, **p < 0.01.

Women

In Model 1 (Table 3), we examined racial/ethnic differences in ADL difficulty for women and found significant racial/ethnic differences in disability. Black and Hispanic women had more ADL difficulty than White women (RR for Black women = 1.89, 95% CI [1.62, 2.21]; RR for Hispanic women = 2.11, 95% CI [1.75, 2.53]). Then, to determine whether physical impairment explained the racial/ethnic disparities, we added impairment to the model (Model 2). Similar to men, impairment was a strong predictor of difficulty with ADL (RR = 1.26, 95% CI [1.23, 1.28]). Unlike in men, racial/ethnic disparities in ADL difficulty among women were not explained by differences in physical impairment, although the effect on race/ethnicity was attenuated. Demographic covariates and socioeconomic and health factors were next added to the model (Model 3). Including SES and chronic burden attenuated the risk ratio for Black (vs. White) women to nonsignificance, but the increased risk in disability for Hispanic women remained significant. Thus, educational attainment, income, and chronic burden explained Black/White differences in ADL difficulty. For all women, higher income was associated with less ADL difficulty, and women with college compared to less than high school education had higher rates of ADL difficulty (Model 3). This is most likely a suppression effect because the association between college education and disability changes direction when chronic burden is added to the model (not shown). Greater chronic burden was associated with higher rates of ADL difficulty as expected.

Table 3.

Rate Ratios for Racial and Ethnic Disparities in Difficulty With Activities of Daily Living as a Function of Individual Factors and the Home Environment, Women (n = 2,835). U.S. National Health and Aging Trends Study (2012).^a

Variable	Model 1: Racial/Ethnic Differences in Disability	Model 2: Physical Impairment (PI)	Model 3: Socioeconomic and Health Factors	Model 4: Indoor Home Environment	Model 5: Impairment and Race/Ethnicity Interactions^e
Race^b
Black	1.89 (1.62, 2.21)**	1.24 (1.06, 1.46)**	1.16 (0.99, 1.36)⁺	1.15 (0.98, 1.35)	1.64 (1.19, 2.26)**
Hispanic	2.11 (1.75, 2.53)**	1.36 (1.14, 1.61)**	1.27 (1.04,1.55)**	1.28 (1.05, 1.56)*	1.58 (1.03, 2.44)*
Other	0.99 (0.68, 1.44)	1.02 (0.65, 1.61)	1.01 (0.63, 1.63)	0.99 (0.63, 1.58)	1.52 (0.78, 2.94)
Age^c
75 to 84	1.71 (1.44, 2.03)**	0.98 (0.82, 1.17)	0.98 (0.81, 1.17)	0.98 (0.82, 1.18)	0.98 (0.82, 1.18)
85+	3.04 (2.58, 3.58)**	1.05 (0.88, 1.26)	1.15 (0.94, 1.40)	1.17 (0.95, 1.43)	1.15 (0.94, 1.41)
Married^d
Divorced			0.89 (0.73, 1.08)	0.87 (0.71, 1.05)	0.86 (0.71, 1.05)
Widowed			0.85 (0.73, 1.00)	0.84 (0.72, 0.98)*	0.84 (0.72, 0.99)*
Never married			0.86 (0.65, 1.14)	0.85 (0.64, 1.13)	0.86 (0.65, 1.14)
PI (SPPB)		1.26 (1.23, 1.28)**	1.23 (1.20, 1.26)**	1.24 (1.21, 1.27)**	1.24 (1.21, 1.27)**
Education ^e
High school			0.93 (0.82, 1.07)	0.93 (0.81, 1.06)	0.92 (0.80, 1.06)
≥College			1.30 (1.06, 1.60)*	1.31 (1.07, 1.61)*	1.31 (1.07, 1.60)*
Income (ln)			0.92 (0.86, 0.99)*	0.93 (0.87, 0.99)*	0.92 (0.87, 0.98)*
Chronic disease			1.29 (1.24, 1.36)**	1.29 (1.24, 1.35)**	1.29 (1.24, 1.35)**
Indoor disrepair present				0.92 (0.72, 1.18)	0.92 (0.71, 1.19)
Indoor clutter present				1.21 (1.05, 1.40)**	1.20 (1.04, 1.39)**
PI × Race/Ethnicity
PI × Black					0.95 (0.92, 0.98)*
PI × Hispanic					0.97 (0.92, 1.02)
PI × Other					0.93 (0.85, 1.02)

Note. SPPB = short physical performance battery.

Significant at *p < 0.05, **p < 0.01, ⁺ p = 0.05.

We examined the interaction between the home environment and physical impairment on disability (Model 4) and found no evidence that the indoor home environment moderated associations between physical impairment and disability or racial/ethnic differences, so the interaction terms were removed from the model. We did find that, like men, women who lived in homes with clutter had a higher rate of ADL difficulty than women in homes without any clutter (RR = 1.21, 95% CI [1.05, 1.40]), although this had no effect on the racial/ethnic disparities in disability. Finally, we tested whether the association between physical impairment and disability varied by race/ethnicity (Model 5) and found that this was, in fact, the case (Figure 2). Hispanic women with no physical impairment had higher rates of ADL difficulty than Whites, but there was no difference in ADL difficulty between Black and White women after accounting for SES and health. However, given the significant interaction between Black race and impairment, at high levels of impairment, there was a significant racial difference in ADL difficulty that remained after accounting for the covariates. Thus, for the most impaired Black women, there are additional factors that may be responsible for their increased level of disability compared to their White counterparts.

Discussion

Identifying individual and contextual factors that contribute to racial/ethnic disparities in disability has the potential to reduce the excess social and economic resources devoted to caring for older disabled individuals in minority communities. Mitigating racial/ethnic disparities in ADLs will lay a foundation for mitigating disparities in mortality. Using nationally representative data on older Americans, we found that racial/ethnic disparities in ADL difficulty were substantial, with Black and Hispanic Americans reporting more difficulty performing ADLs than Whites. The disparities were fully explained by physical impairment in men, but among women, physical impairment did not explain the large racial/ethnic disparities in ADL difficulty.

Increased clutter in the home was associated with higher rates of disability, and associations of physical impairment and disability were stronger for men living in more cluttered home environments. Although the home environment was an important risk factor for disability, it did not explain racial/ethnic disparities in disability for women. Accounting for income, education, and chronic burden eliminated the racial disparities for Black women but did not explain the disparities for Hispanic women. Lastly, we found that for Black women with moderate to high levels of physical impairment, higher rates of disability remained even after accounting for other risk factors.

Our results confirm previous research that identifies strong racial and ethnic disparities in physical disability among older Americans (Fuller-Thomson et al., 2009; Mendes de Leon, et al., 2005; Thorpe, Weiss, Xue, & Fried, 2009; Warner & Brown, 2011). We expand on this research by including measures of the indoor home environment to determine whether the home context contributes to the racial/ethnic disparities beyond SES and health. Additionally, we considered models of racial/ethnic disabilities separately by gender, which may help identify the sources of the remaining unexplained disparities in disability for women. In line with our hypotheses, differences in physical impairment were not responsible for racial/ethnic disparities in ADL difficulty for women despite explaining differences for men. Although the Black/White differences were fully explained by SES and health for women with no physical impairment, the racial disparity remained at moderate and high levels of impairment, suggesting that unmeasured factors are responsible for the disparity in Black women with worse physical functioning. Contrary to our hypotheses, clutter and disrepair were not responsible for racial/ethnic disparities in ADL difficulty among women. It is possible that for older Hispanic and lower functioning Black women, a lifetime of exposure to stressors results in more severe disability that is not evident for Black men. This is consistent with the weathering hypothesis, which posits accelerated aging and cumulative wear on the health of African American women, as they are exposed to multiple sources of social and economic inequality over the life course (Geronimus, 1991). Differences in disability for older non-White women may be attributable to a life history shaped by both their gender and race, as outlined in intersectionality theory (Schulz & Mullings, 2006). These results are in line with previous research (Fuller-Thomson et al., 2009; Mendes de Leon et al., 2005; Warner & Brown, 2011) finding greater disability for Black women compared to Black men or White women, which is not explained by SES. Hispanic women had the highest degree of disability in our sample compared to White women, and this ethnic disparity in disability was not explained by the factors we examined in our analysis. Researchers have determined that older Hispanic adults are less likely to go into nursing homes for care compared to Whites, and this may result in a more disabled community-living Hispanic community (Thomeer, Mudrazija, & Angel, 2015). Additional investigation of factors that influence difficulty with self-care in the Hispanic community could assist in determining how to most effectively reduce disability for older Hispanics based on their preferences, culture, and structural barriers to care.

While we found racial/ethnic differences in disability for women, they varied across levels of physical impairment. Although older Black women reported similar ADL difficulty to White women, a racial gap in disability increased with increasing physical impairment. Researchers have not yet examined reasons for differences in the effect of physical capacity on disability by race, but our results suggest that it is critical to begin to consider life-course exposures that reflect the dual identity of being Black and female, which may contribute to higher rates of disability in Black women as they age. Black women hold two low-status positions that expose them to unique forms of oppression and discrimination over their life course to influence their health later in life (Schulz & Mullings, 2006). Thus, racial/ethnic differences in disability for women may be established earlier in life and not a function of more immediate factors like physical impairment or income. Researchers have found, for instance, that changes in disability are associated with a sense of mastery (Kempen, van Sonderen, & Ormel, 1999), and these psychological attributes may be related to earlier life experiences shaped by race/ethnicity and gender. Future research could include earlier life measures of stressors and discrimination, which may account for some of the racial/ethnic disparities in disability among older women.

We considered the home as a critical feature of the environment, which may interact with physical impairment to influence disability. We found that physical barriers in the form of clutter in the home are associated with difficulty performing basic daily tasks like dressing and bathing. Given the relative ease with which these barriers can be addressed to reduce disability, a simple intervention such as helping older adults remove clutter from their house is likely to be efficacious and effective. Our results suggest that home modifications might be most helpful to improve ADL disability for men with the highest degree of physical impairment.

It is possible that the neighborhood environment, which was not considered in this study, may be associated with racial/ethnic disparities in disability in women. Researchers have identified multiple dimensions of the neighborhood that may influence disability including: SES, traffic, noise, lighting, street connectivity, land use, and crime (Beard et al., 2009; Clarke, Ailshire, Bader, Morenoff, & House, 2008; Clarke & George, 2005; Freedman, Grafova, Schoeni, & Rogowski, 2008; Satariano et al., 2014). Although these factors have not been examined in relation to racial/ethnic disparities in disability, this would be important for future research given the strong racial/ethnic patterning of neighborhoods (Massey & Denton, 1993) and evidence that neighborhood context is associated with racial/ethnic disparities in health (King, Morenoff, & House, 2011; Morenoff et al., 2007). Additionally, neighborhood attributes such as residential segregation and socioeconomic disadvantage may better reflect chronic exposure to risks and access to resources over the life course that are related to the intersection of race/ethnicity and gender and also associated with late life disability.

This study was not without limitations, including the cross-sectional analysis. Our results cannot be interpreted as causal. However, given that measures of SES in an older population most likely reflect a state earlier in adulthood, we have increased confidence that the socioeconomic exposures preceded disablement. While we found that living in a more cluttered home was associated with greater risk of self-care disability, it is not possible to determine from our cross-sectional analysis whether clutter in the home results in more disability or whether individuals with more disability are unable to care for their homes, resulting in more clutter (Gitlin, et al., 2001). Additionally, our results may be biased by people who did not survive long enough to participate in the study. Survivor bias is likely, given the older average age of our sample at baseline, and would likely result in an underestimation of the racial/ethnic disparities in disability (Kelley-Moore & Ferraro, 2004). Similarly, study attrition may bias the results, as individuals with greater disability are more likely to drop out of the study. Although a variety of additional environmental, social, and psychological factors may be associated with racial/ethnic disparities in disability, NHATS is somewhat limited in psychosocial measures to assess stress, discrimination, and coping. In a post hoc analysis, we added a series of cognitive and psychosocial predictors to our model to determine whether they might be responsible for the remaining racial/ethnic disparities in ADL difficulty for women. Although having more depressive and anxiety symptoms was associated with higher rates of ADL difficulty and greater social participation was associated with less difficulty (not shown), accounting for these factors did not affect the significance or magnitude of other predictors in the model so they were not included.

One of the greatest attributes of the NHATS is the assessment of disability and disability-related factors. The measure of ADL difficulty we used extends beyond traditional measures of disability by incorporating assistive device usage and assistance from others (Freedman et al., 2011). By using a more comprehensive measure of ADL, our results may represent a more accurate description of racial/ethnic disparities in disability. Other researchers have found, for instance, that Blacks and Hispanics have higher rates of mobility device usage than Whites (Cornman & Freedman, 2008). These differences were accounted for by our measure of ADL difficulty, which suggests that racial/ethnic disparities in disability persist even after accounting for differences in device usage.

Conclusions

Our results suggest substantial differences in the factors that contribute to racial/ethnic disparities in disability for older women versus older men. This calls for developing models of disability that account for the different life experiences that are unique to not only women versus men but to Black and Hispanic women more specifically. There is little evidence that we can mitigate racial/ethnic disparities in disability for women by addressing socioeconomic, medical, behavioral, and home environmental causes. By applying a more comprehensive approach to studying racial/ethnic disparities in disability and accounting for exposures over the life course, it may be possible to identify the causes for racial/ethnic disparities in disability for women as well as factors on which we can intervene to reduce and eventually eliminate disparities in later life.

Footnotes

Authors’ Note

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging or the National Institutes of Health.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research is supported by the National Institutes of Health National Institute on Aging (R03 AG043661-02). The National Health and Aging Trends Study (NHATS) is sponsored by the National Institute on Aging (Grant number NIA U01AG032947) through a cooperative agreement with the Johns Hopkins Bloomberg School of Public Health.

Note

References

Beard

J. R.

Blaney

Cerda

Frye

Lovasi

G. S.

Ompad

… Vlahov

(2009). Neighborhood characteristics and disability in older adults. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 64B, 252–257. doi:10.1093/geronb/gbn018

Bowen

M. E.

(2009). Childhood socioeconomic status and racial differences in disability: evidence from the Health and Retirement Study (1998–2006). Social Science & Medicine, 69, 433–441.

Clarke

Ailshire

J. A.

Bader

Morenoff

J. D.

House

J. S.

(2008). Mobility disability and the urban built environment. American Journal of Epidemiology, 168, 506–513. doi:http://dx.doi.org/10.1093/aje/kwn185

Clarke

George

L. K.

(2005). The role of the built environment in the disablement process. American Journal of Public Health, 95, 1933–1939.

Cornman

J. C.

Freedman

V. A.

(2008). Racial and ethnic disparities in mobility device use in late life. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 63, S34–S41. doi:http://dx.doi.org/10.1093/geronb/63.1.S34

Dunlop

D. D.

Song

Manheim

L. M.

Daviglus

M. L.

Chang

R. W.

(2007). Racial/ethnic differences in the development of disability among older adults. American Journal of Public Health, 97, 2209–2215. doi:http://dx.doi.org/10.2105/AJPH.2006.106047

Erickson

Lee

von Schrader

. (2010). 2008 Disability Status Report: the United States. Ithaca, NY: Cornell University Rehabilitation Research and Training Center on Disability Demographics and Statistics.

Freedman

V. A.

(2009). Adopting the ICF language for studying late-life disability: A field of dreams? The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 64A, 1172–1174. doi:10.1093/gerona/glp095

Freedman

V. A.

Grafova

I. B.

Schoeni

R. F.

Rogowski

(2008). Neighborhoods and disability in later life. Social Science & Medicine, 66, 2253–2267. doi:http://dx.doi.org/10.1016/j.socscimed.2008.01.013

10.

Freedman

V. A.

Kasper

J. D.

Cornman

J. C.

Agree

E. M.

Bandeen-Roche

Mor

… Wolf

D. A.

(2011). Validation of new measures of disability and functioning in the national health and aging trends study. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 66A, 1013–1021. doi:10.1093/gerona/glr087

11.

Fried

L. P.

Guralnik

J. M.

(1997). Disability in older adults: Evidence regarding significance, etiology, and risk. Journal of the American Geriatrics Society, 45, 92–100. doi:10.1111/j.1532-5415.1997.tb00986.x

12.

Fuller-Thomson

Nuru-Jeter

Minkler

Guralnik

J. M.

(2009). Black-White disparities in disability among older Americans: Further untangling the role of race and socioeconomic status. Journal of Aging and Health, 21, 677–698. doi:http://dx.doi.org/10.1177/0898264309338296

13.

Geronimus

A. T.

(1991). The weathering hypothesis and the health of African-American women and infants: Evidence and speculations. Ethnicity & Disease, 2, 207–221.

14.

Gitlin

L. N.

(2003). Conducting research on home environments: Lessons learned and new directions. The Gerontologist, 43, 628–637. doi:10.1093/geront/43.5.628

15.

Gitlin

L. N.

Mann

Machiko

Sue

M. M.

(2001). Factors associated with home environmental problems among community-living older people. Disability & Rehabilitation, 23, 777–787.

16.

Guralnik

J. M.

Simonsick

E. M.

Ferrucci

Glynn

R. J.

Berkman

L. F.

Blazer

D. G.

… Wallace

R. B.

(1994). A short physical performance battery assessing lower extremity function: Association with self-reported disability and prediction of mortality and nursing home admission. Journal of Gerontology, 49, M85–M94. doi:http://dx.doi.org/10.1093/geronj/49.2.M85

17.

Haas

Rohlfsen

(2010). Life course determinants of racial and ethnic disparities in functional health trajectories. Social Science & Medicine, 70, 240–250. doi:http://dx.doi.org/10.1016/j.socscimed.2009.10.003

18.

Haas

S. A.

Krueger

P. M.

Rohlfsen

(2012). Race/Ethnic and nativity disparities in later life physical performance: The role of health and socioeconomic status over the life course. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 67B, 238–248. doi:10.1093/geronb/gbr155

19.

Harper

(2014). Economic and social implications of aging societies. Science, 346, 587–591.

20.

Iwarsson

Isacsson

Å.

Lanke

(1998). ADL dependence in the elderly population living in the community: The influence of functional limitations and physical environmental demand. Occupational Therapy International, 5, 173–193. doi:10.1002/oti.74

21.

Iwarsson

Wahl

H.-W.

Nygren

Oswald

Sixsmith

… Tomsone

(2007). Importance of the home environment for healthy aging: Conceptual and methodological background of the European ENABLE–AGE Project. The Gerontologist, 47, 78–84. doi:10.1093/geront/47.1.78

22.

Johns Hopkins University. (2011). National Health and Aging Trends Study (NHATS). Retrieved from January 2, 2015, NHATS, https://www.nhatsdata.org/

23.

Kelley-Moore

J. A.

Ferraro

K. F.

(2004). The black/white disability gap: persistent inequality in later life? The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 59, S34–S43. doi:http://dx.doi.org/10.1093/geronb/59.1.S34

24.

Kempen

G. I.

van Sonderen

Ormel

(1999). The impact of psychological attributes on changes in disability among low-functioning older persons. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 54, P23–P29. doi:http://dx.doi.org/10.1093/geronb/54B.1.P23

25.

King

K. E.

Morenoff

J. D.

House

J. S.

(2011). Neighborhood context and social disparities in cumulative biological risk factors. Psychosomatic Medicine, 73, 572–579.

26.

Kington

R. S.

Smith

J. P.

(1997). Socioeconomic status and racial and ethnic differences in functional status associated with chronic diseases. American Journal of Public Health, 87, 805–810.

27.

Latham

(2012). Progressive and accelerated disability onset by race/ethnicity and education among late midlife and older adults. Journal of Aging and Health, 24, 1320–1345. doi:http://dx.doi.org/10.1177/0898264312459345

28.

Latham

(2014). Racial and educational disparities in mobility limitation among older women: What is the role of modifiable risk factors? The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 69, 772–783. doi:10.1093/geronb/gbu028

29.

Massey

D. S.

Denton

N. A.

(1993). American Apartheid: Segregation and the making of the underclass. Cambridge, MA: Harvard University Press.

30.

Mendes de Leon

C. F.

Barnes

L. L.

Bienias

J. L.

Skarupski

K. A.

Evans

D. A.

(2005). Racial disparities in disability: Recent evidence from self-reported and performance-based disability measures in a population-based study of older adults. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 60, S263–S271. doi:10.1093/geronb/60.5.S263

31.

Mendes de Leon

C. F. M.

Beckett

L. A.

Fillenbaum

G. G.

Brock

D. B.

Branch

L. G.

Evans

D. A.

Berkman

L. F.

(1997). Black-white differences in risk of becoming disabled and recovering from disability in old age: A longitudinal analysis of two EPESE populations. American Journal of Epidemiology, 145, 488–497. doi:http://dx.doi.org/10.1093/oxfordjournals.aje.a009136

32.

Metzelthin

S. F.

van Rossum

de Witte

L. P.

Ambergen

A. W.

Hobma

S. O.

Sipers

Kempen

G. I. J. M.

(2013). Effectiveness of interdisciplinary primary care approach to reduce disability in community dwelling frail older people: Cluster randomised controlled trial. British Medical Journal, 347. doi:10.1136/bmj.f5264

33.

Minkler

Fuller-Thomson

Guralnik

J. M.

(2006). Gradient of disability across the socioeconomic spectrum in the United States. New England Journal of Medicine, 355, 695–703. doi:10.1056/NEJMsa044316

34.

Montaquila

Freedman

V. A.

Kasper

J. D.

(2012). National Health and Aging Trends Study Round 1 Income Imputation: Johns Hopkins University School of Public Health.

35.

Montaquila

Freedman

V. A.

Spillman

B. C.

Kasper

J. D.

(2014). National health and aging trends study development of round 2 survey weights. Baltimore, MD: Johns Hopkins School of Public Health.

36.

Morenoff

J. D.

House

J. S.

Hansen

B. B.

Williams

D. R.

Kaplan

G. A.

Hunte

H. E.

(2007). Understanding social disparities in hypertension prevalence, awareness, treatment, and control: The role of neighborhood context. Social Science & Medicine, 65, 1853–1866.

37.

Murtagh

K. N.

Hubert

H. B.

(2004). Gender differences in physical disability among an elderly cohort. American Journal of Public Health, 94, 1406–1411. doi:http://dx.doi.org/10.2105/AJPH.94.8.1406

38.

Nachreiner

N. M.

Findorff

M. J.

Wyman

J. F.

McCarthy

T. C.

(2007). Circumstances and consequences of falls in community-dwelling older women. Journal of Women’s Health, 16, 1437–1446. doi:10.1089/jwh.2006.0245

39.

Nagi

S. Z.

(1965). Some conceptual issues in disability and rehabilitation. Washington, DC: American Sociological Association.

40.

Penning

M. J.

Strain

L. A.

(1994). Gender differences in disability, assistance, and subjective well-being in later life. Journal of Gerontology, 49, S202–S208.

41.

Pinquart

Sörensen

(2005). Ethnic differences in stressors, resources, and psychological outcomes of family caregiving: A meta-analysis. The Gerontologist, 45, 90–106. doi:10.1093/geront/45.1.90

42.

Satariano

Kealey

Hubbard

Kurtovich

Ivey

Bayles

… Prohaska

(2014). Mobility disability in older adults: At the intersection of people and places. The Gerontologist. doi:10.1093/geront/gnu094

43.

Schoeni

R. F.

Martin

L. G.

Andreski

P. M.

Freedman

V. A.

(2005). Persistent and growing socioeconomic disparities in disability among the elderly: 1982–2002. American Journal of Public Health, 95, 2065–2070. doi:10.2105/ajph.2004.048744

44.

Schulz

A. J.

Mullings

(2006). Gender, race, class, and health: Intersectional approaches. San Francisco, CA: Jossey-Bass.

45.

Seeman

T. E.

Merkin

S. S.

Crimmins

E. M.

Karlamangla

A. S.

(2010). Disability trends among older Americans: National health and nutrition examination surveys, 1988-1994 and 1999-2004. American Journal of Public Health, 100, 100–107. doi:http://dx.doi.org/10.2105/AJPH.2008.157388

46.

StataCorp. (2015). Stata Statistical Software: Release 14. College Station, TX: Author.

47.

Szanton

S. L.

Wolff

J. L.

Leff

Roberts

Thorpe

R. J.

Tanner

E. K.

… Bishai

(2015). Preliminary data from community aging in place, advancing better living for elders, a patient-directed, team-based intervention to improve physical function and decrease nursing home utilization: The first 100 individuals to complete a Centers for Medicare and Medicaid Services Innovation Project. Journal of the American Geriatrics Society, 63, 371–374. doi:10.1111/jgs.13245

48.

Tabbarah

Mihelic

Crimmins

E. M.

(2001). Disability: The demographics of physical functioning and home environments of older Americans. Journal of Architectural and Planning Research, 18, 183–193.

49.

Thomeer

M. B.

Mudrazija

Angel

J. L.

(2015). How do race and Hispanic ethnicity affect nursing home admission? Evidence from the health and retirement study. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 70, 628–638. doi:10.1093/geronb/gbu114

50.

Thorpe

R. J.

Kasper

J. D.

Szanton

S. L.

Frick

K. D.

Fried

L. P.

Simonsick

E. M.

(2008). Relationship of race and poverty to lower extremity function and decline: Findings from the women’s health and aging study. Social Science & Medicine, 66, 811–821. doi:http://dx.doi.org/10.1016/j.socscimed.2007.11.005

51.

Thorpe

R. J.

Koster

Kritchevsky

S. B.

Newman

A. B.

Harris

Ayonayon

H. N.

… Study

B. C.

(2011). Race, socioeconomic resources, and late-life mobility and decline: Findings from the health, aging, and body composition study. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 66A, 1114–1123. doi:10.1093/gerona/glr102

52.

Thorpe

R. J.

Weiss

Xue

Q.-L.

Fried

(2009). Transitions among disability levels or death in African American and White older women. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 64A, 670–674. doi:10.1093/gerona/glp010

53.

Vasunilashorn

Coppin

A. K.

Patel

K. V.

Lauretani

Ferrucci

Bandinelli

Guralnik

J. M.

(2009). Use of the short physical performance battery score to predict loss of ability to walk 400 meters: Analysis from the InCHIANTI study. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 1–7. doi:http://dx.doi.org/10.1093/gerona/gln022

54.

Verbrugge

L. M.

Jette

A. M.

(1994). The disablement process. Social Science & Medicine, 38, 1–14. doi:http://dx.doi.org/10.1016/0277-9536(94)90294-1

55.

Wahl

H.-W.

Fänge

Oswald

Gitlin

L. N.

Iwarsson

(2009). The home environment and disability-related outcomes in aging individuals: What is the empirical evidence? The Gerontologist, 49, 355–367.

56.

Wahl

H.-W.

Schilling

Oswald

Iwarsson

(2009). The home environment and quality of life-related outcomes in advanced old age: Findings of the ENABLE-AGE project. European Journal of Ageing, 6, 101–111. doi:10.1007/s10433-009-0114-z

57.

Wan

Larsen

L. J.

(2014). Older Americans with a disability: 2008–2012 American Community Survey Reports (Vol. ACS-29). Washington, DC: U.S. Census Bureau.

58.

Warner

D. F.

Brown

T. H.

(2011). Understanding how race/ethnicity and gender define age-trajectories of disability: An intersectionality approach. Social Science & Medicine, 72, 1236–1248. doi:http://dx.doi.org/10.1016/j.socscimed.2011.02.034

59.

White

I. R.

Royston

Wood

A. M.

(2011). Multiple imputation using chained equations: Issues and guidance for practice. Statistics in Medicine, 30, 377–399. doi: 0.1002/sim.4067

60.

Williams

D. R.

Priest

Anderson

(2016). Understanding associations between race, socioeconomic status and health: Patterns and prospects. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association, 35, 407–411. doi:10.1037/hea0000242

61.

World Health Organization. (2002). International classification of functioning, disability and health: ICF. Geneva, Switzerland: Author.

62.

Zsembik

B. A.

Peek

M. K.

Peek

C. W.

(2000). Race and ethnic variation in the disablement process. Journal of Aging and Health, 12, 229–249. doi:http://dx.doi.org/10.1177/089826430001200205